A survey on advances and insights of image analysis techniques for phenotyping in maize research: systematic review

Maize (Zea mays L), a cultigen derived from domesticated teosinte, is the queen of cereals because of its broad environmental adaptability and high genetic yield potential. It has major economic importance as both raw corn and as feedstock for numerous value-added products, with each kernel type serving multiple agri-industries. Driven by technological advances and expanded resources, breeders and researchers increasingly prioritize maize breeding and development. This paper systematically reviews the last decade of image analysis advancements in maize, focusing on techniques adopted for phenotyping, plant classification and disease identification. We examine why image-based phenotyping is necessary for modern agriculture, why maize is a primary model for image analysis and which maize traits remain underexplored. We summarize how machine learning and deep learning simplify image processing and feature extraction; identify the main challenges researchers encounter when adopting these methods and propose potential solutions. Our review shows that integrating advanced imaging, computer vision and AI have enabled earlier stress detection, improved yield prediction and more efficient trait mapping in maize. Continued innovation in scalability, robustness and interpretability is essential to translate these technological advances into real-world agricultural impact.